Eigenvalue-based cooperative spectrum sensing, because of its robustness, has attracted a lot of attention. Computational complexity is a major drawback of this method. In this paper, to improve the detection probability and to reduce the number of affective secondary users, we investigate the effect of secondary users different distances from primary user base station with a partial clustering of secondary users. Monte-Carlo simulation results show the effectiveness of proposed method specially in low SNR values compared to other methods.